three-man-team
A structured 3-agent AI dev team — Architect, Builder, Reviewer. Built from production use. Token-optimized. Works with Claude Code, VS Code, Cursor, and any AI
What is three-man-team?
A structured 3-agent AI dev team — Architect, Builder, Reviewer. Built from production use. Token-optimized. Works with Claude Code, VS Code, Cursor, and any AI that supports context files.
Overview of three-man-team
What it is
Three Man Team is a structured 3-agent AI development team — Architect, Builder, and Reviewer. It is built from production use, token-optimized, and works with Claude Code, VS Code, Cursor, and any AI that supports context files. Created by Russell Aaron.
Key features
- Three distinct agents with clear roles: Architect (plans and deploys), Builder (builds exactly to the brief), Reviewer (catches mistakes).
- A defined workflow: every unit of work follows Architect → Builder → Reviewer → Architect deploy path.
- Token optimization: five built-in rules in CLAUDE.md that reduce unnecessary tool calls and output, auto-loaded every session.
- Auto-update: Arch checks a version registry at session start and prompts for critical or non-critical updates.
- Two install modes: per-project (recommended) and global.
- Templates provided:
project-folder(pre-configured personas) andgeneric(blank slate with placeholders).
Use cases
Three Man Team addresses the lack of discipline in AI coding tools. It is suitable for any software development project that benefits from a structured process of planning, building, and reviewing. It works especially well with Claude Code, VS Code, Cursor, and other AI tools that support context files.
Getting started
Per-project install (recommended): bash git clone https://github.com/russelleNVy/three-man-team.git .claude/skills/three-man-team cd .claude/skills/three-man-team && ./setup For global install and further instructions, see the repository README.
License and pricing
License: MIT. Free forever.
Based on the project's public README and repository metadata.